For many reason, the flatfile decoding/loading against the table definition can fail. To help, you could still add this property into EBCDIC loading step: debug.folder=data/Debug
It enables to declare a folder into which a CSV file is created corresponding to the stream pass through the COPY statement of POSTGRESQL. It enables to check eventual mistake.
A log file is also create indicating how the ebcdic file has been cut out to fill each column.
Example:
Flat file
debug.csv file
cut off log file